spm.impute: Multiple Data Imputation with SPM

In izhbannikov/spm: Stochastic Process Model for Analysis of Longitudinal and Time-to-Event Outcomes

Multiple Data Imputation with SPM

Description

Multiple Data Imputation with SPM

Usage

spm.impute(
  x,
  id = 1,
  case = 2,
  t1 = 3,
  t2 = 3,
  covariates = 4,
  minp = 5,
  theta_range = seq(0.01, 0.2, by = 0.001)
)

Arguments

x	A longitudinal dataset with missing observations
id	A name (text) or index (numeric) of ID column. Default: 1
case	A case status column name (text) or index (numeric). Default: 2
t1	A t1 (or t if short format is used) column name (text) or index (numeric). Default: 3
t2	A t2 column name (if long format is used) (text) or index (numeric). Default: 4
covariates	A list of covariate column names or indices. Default: 5
minp	Number of imputations. Default: 5
theta_range	A range of parameter theta used for optimization, default: seq(0.01, 0.15, by=0.001).

Value

A list(imputed, imputations)

imputed An imputed dataset.

imputations Temporary imputed datasets used in multiple imputaitons.

Examples

## Not run: 
library(stpm) 
##Data preparation ##
data <- simdata_discr(N=1000, dt = 2)
miss.id <- sample(x=dim(data)[1], size=round(dim(data)[1]/4)) # ~25% missing data
incomplete.data <- data
incomplete.data[miss.id,5] <- NA
incomplete.data[miss.id-1,6] <- NA
## End of data preparation ##

# Estimate parameters from the complete dataset #
p <- spm_discrete(data, theta_range = seq(0.075, 0.09, by=0.001))
p

##### Multiple imputation with SPM #####
imp.data <- spm.impute(x=incomplete.data, 
                      minp=5, 
                      theta_range=seq(0.075, 0.09, by=0.001))$imputed
head(imp.data)
## Estimate SPM parameters from imputed data and compare them to the p ##
pp.test <- spm_discrete(imp.data, theta_range = seq(0.075, 0.09, by=0.001))
pp.test

## End(Not run)

izhbannikov/spm documentation built on Sept. 10, 2022, 12:15 p.m.